Minutes of Forward Tracker Upgrade Meeting on Thursday 10/12/98 =============================================================== Present: Steve Burke, Tim Greenshaw, Rob Henderson, Derrick Hill, Graham Houghton, Jim Kay, John Morris, Girish Patel, Dave Sankey, Graham Stokes. 1) Actions from Previous Meeting. --------------------------------- Tim --- showed the updated project budget. On the day following the meeting this was further corrected on receipt of an offer for double-sided PCB manufacture from Spemco (the offer shown in the meeting was for single-sided PCB only, which explains why it was so much less than was expected). Allowing 10k pounds for the trigger (roughly enough to equip 16 channels with APDs) and 2 pounds per pin, we have about 30k pounds ``uncommitted'' - including all major items of which we are aware. Tim also discussed the results of a visual inspection of the test PCBs. These look good with the exception of a bulge near some of the plate-through holes. John Morton tested the PCBs in dry air. Between adjacent tracks potential differences of 3kV could be supported. It was noted that the PCBs tested were tinned. Those used in the chambers must be neither tinned nor lacquered. Action. Graham to discuss features near plate-throughs with Spemco and to ensure that PCBs not tinned/lacquered. Graham H -------- reported that worries about the capacity of the dowels to carry the load of 8 wires were unfounded. The dowels actually sit in a more favourable location in the thickened Noryl end pieces (no gap between Noryl and Nomex at the dowel position) and calculations show there is a large safety margin. Graham also showed the pin design for the 8 wire chambers. This is similar to that for the current planars with the additional length being in the loose fitting section within the Noryl end piece. The pin is simplified in that the outer diameter of the section which fits into the connector has been made the same as the outer diameter of the loose fitting section in the Noryl, simplifying machining. Photos and drawings show that the tank feet can be offset rearwards by 6cm as long as the (MWPC) feedthroughs adjacent to the feet are not used. A possible design for the feet was shown. As emergency fall-back solutions it would be possible to re-use the existing feet on the central U channel, or the offset feet on the opposite sides for which they are being designed to shift the load point forward along the rail, rather than backward. These latter solutions would leave a large unsupported overhanging section of the tamk and are to be avoided if possible! It was agreed that ee should construct feet that will allow us a little room for manouevre (~7cm forward shift of FT?) if possible! Action. Tim to confirm with Manfred that 6cm is all that is required. Derrick ------- showed us the four Medelec pins which have arrived. These appeared to be made entirely of copper (no brass outer or stainless steel collar). During the discussion on the pins it was suggested that the pin design could be simplified still further by making the outer diameter of the section inside the Noryl constant and drilling out the Noryl to leave a section with the necessary gap between Noryl and pin. In a later discussion, Graham explained that this idea does not work as it requires that the full length of the section of the pin in the Noryl be pushed through the very tight fit during the assembly procedure, with the attendant risk of damaging the pin ot the Noryl. Given the difficulties in obtaining the pins it was agreed that we need to find alternative suppliers. There was some discussion on the tolerances for the pin positions. It was pointed out that 50 micron tolerances are about the best that can be reasonably achieved using the proposed machining techniques. It was felt that 10 micron error (standard deviation) on the pins transverse positions w.r.t one another would be desirable, lower tolerances were acceptable in the z coordinate. (Transverse position means at 90 degrees to the wire, but parallel to the long side of the cell x-section; z coordinate means at 90 degrees to the wire but parallel to the short side of the cell x-sect, i.e. these are the transverse and z directions in the H1 coordinate system.) No-one knew what the relationship between ``tolerance'' and standard deviation is. Action. Derrick to investigate other UK suppliers of pins. Don to inquire about pin manufacture in Germany (Bishoff). Tim to find out what tolerances were for machining radial Noryl and to clarify relationship between tolerance and standard deviation. Dave ---- explained that the pre-amp test system was now functioning perfectly. The available 5 pre-amps have been tested and appear to be OK, though there was some concern that some features, such as rise-times, were at the boundaries specified as acceptable. The remaining 95 pre-amps have now been ordered. Girish ------ reported that he now has the 6 wire planar simulation going at RAL. It will take a little time to change this to an 8 wire version. He showed us the results of processing high Q^2 data with all current chambers on, with the radials removed, with the rearmost 4 wires in each radial off (those closest to the window), with the two rearmost and two foremost radial wires off and with the 4 foremost radial wires off. There were no dramatic changes in the numbers of tracks found. The most marked change was an increase in the quality of the fit for combinations which removed the rearmost two radial wires. This suggests that the drift velocities for these two wires may be incorrect. They are currently assumed to be the same as those for all other radial wires. Given the umpleasant field configuration near the rearmost wires, this is perhaps not surprising. Action ------ Girish 8 wire planars in simulation, get GO simulation going. 2) Extra Walls. --------------- It was decided to order enough PCBs and Nomex panels to construct two more 8 wire planar chambers (i.e. a total of 6, one for learning construction procedures on, the 3 planned for installation during the HERA upgrade and 2 in case it should prove necessary to replace the radial chambers). 3) New planars, 6 or 8 wire? ---------------------------- It was decided to build 8 wire planar chambers. 5) Status of software development. ---------------------------------- John gave an overview of the FT Upgrade software situation. It was noted that the changes discussed above (6 to 8 wire planars) will cause some delay, but this was thought to be a small effect. John stressed that the following topics are still not covered and that additional manpower is required for: a) Monitoring and Calibration. b) Alignment. c) Event display. Rob expressed his concern about the alignment situation and agreed to investigate what can be done at Lancaster to produce an independent means of monitoring and measureing the planar drift velocity. Action Rob to find out if drift velocity monitoring can be studied at Lancaster. 6) Year 2000 compliance. ------------------------ Tim informed everyone that we have been officially warned that all our software must be Y2k compliant! 7) Next Meeting. ---------------- The next meeting is scheduled for 28/1/99. This can not be at DESY as suggested as Tim has to organise the UCAS visit day at Liverpool on 27/1/99. How about RAL for a change? 8) AoB. ------- I wish all involved with the project a Merry Christmas and a Peaceful and Prosperous New Year! Tim.